I think peer review and the openess of information involving the experiments does alot to prevent a bias incorrect result from being found as correct.